Replacing words does not create a sound paraphrase
A paraphrase presents another person's idea in a new form suited to your argument. Synonyms inside the original syntax retain its structure. Yet a highly free rewrite can distort the claim. Independence and fidelity therefore have to be assessed together rather than treated as alternatives.
The source still needs attribution when every word has changed. Citation recognises data, argument, distinction or interpretation as well as exact wording. Keep general knowledge and your own inference visibly separate from that borrowed contribution.
The worksheet is citation-style neutral. It does not decide whether a programme requires footnotes, author-date references or page numbers. It exposes content quality before the final formatting check.
Step one: compare the complete meaning
Identify the actor or object, relationship or action, result and every important qualification in the original. Check each against the paraphrase. Negation, uncertainty and contrast disappear easily. “May be associated” cannot become “causes”, and “no evidence was found” does not mean “was disproved”.
Resolve pronouns and references such as “this group”, “there” or “under that condition”. They may point to earlier sentences, making the isolated extract insufficient. Read the surrounding passage and state the actual referent.
Technical terms may remain where substitution would reduce accuracy. Do not force obscure synonyms merely to avoid every repeated word. The objective is a fresh conceptual account, not artificial vocabulary.
Step two: construct an independent form
Close the source and explain the idea from understanding. Change perspective, information order and sentence boundaries only where the meaning survives. Reopen the original and mark distinctive sequences, matching lists and the same rhetorical rhythm.
A patchwork of small fragments from different passages remains unsafe. If a memorable expression is necessary or itself under analysis, quote it directly and mark additions or omissions according to the required convention.
Translation is not automatic originality. An idea drawn from another-language source still needs attribution. Where exact language matters, indicate that a translation is yours and preserve access to the source wording.
Step three: limit scope and argumentative function
Compare population, time, location, material and method. A case report does not establish a universal rule. Correlation does not establish cause, a theoretical proposal is not a measured effect, and a sentence in the discussion does not automatically report primary data.
Inspect the function in your paragraph too. Has an illustrative example become the central proof? Is the source supplying a definition, finding, criticism or contextual opening? A linguistically accurate paraphrase can still mislead when deployed for a different argumentative task.
Write the strongest permissible interpretation beside an impermissible stronger one. This contrast quickly exposes hidden escalation.
Step four: assign the source locally and unambiguously
Place the reference close enough to show what it supports. One citation at the end of a long paragraph may not distinguish several borrowed and original propositions. Repeat or distribute citations where the attribution would otherwise shift.
Check author, year, title or short form and precise location under the required style. For an unpaginated web source, a section or heading may help if the style permits it. Confirm that every short citation resolves to one entry in the reference list.
Do not present a primary work as personally consulted when you only read its treatment in a secondary source. Use the prescribed secondary-citation convention and limit the claim to what you actually verified.
Build checking into the drafting cycle
Do not leave the entire check until the final day. Mark new paraphrases in the working draft and review them section by section with the source open. Record a simple outcome: released, changed to quotation, narrowed or removed.
A second reader may expose a boundary case but cannot replace your source comparison. Similarity software cannot reliably identify every inflated claim or a fully reworded but unattributed idea.
The source-note template preserves origin, while the quotation inventory checks the finished draft. Browse the research methods hub; the one foundation route is how verifiable checking works.
For a final sample, select one paraphrase from the introduction, method, analysis and discussion. Apply all four steps without relying on memory. If the same failure recurs, inspect the surrounding notes rather than fixing only the sampled sentence. A repeated scope error often signals that source summaries were too broad from the start.